Widok Energy Cultures in the European Union

Remigiusz ROSICKI

Adam Mickiewicz University in Poznań

DOI 10.14746/ps.2016.1.17

ENERGY CULTURES IN THE EUROPEAN UNION

The present text aims to analyse “energy cultures” in the European Union, and thus attempts to verify the assumption underlying the legitimacy of statements positing a di­ vision of the European Union member states according to particular kinds of energy consumption practice. In the first place one should point to two main directions of en­ ergy cultures analysis, which can be presented as follows: (1) a presentation of a culture in the form of the conversion of resources as well as the impact of the conversion on the natural environment, (2) an energy culture as a special sphere of social awareness. These two directions of thought are connected with the two main roots of research into culture itself: (1) culture as farming and animal husbandry (presently the domain ofnat- ural sciences), and (2) culture as a body of immaterial artefacts and a reproduction of social practice (presently the domain o f social sciences and humanities) (Kłosowska, 1969; Kłosowska, 1972; Burszta, 1998: 35-57; Strinati, 1998: 15-49).

It should be stressed that in a great many cases - despite invoking “humanistic strat­ egies in the attitude towards energy issues”, that is to the approaches to the energy cul­ ture as a specific kind of public awareness - the analyses are frequently nothing but the issues concerned with data such as the GHG emission index, the energy intensity index, the quantitative data on primary energy production, etc. (Łucki, Misiak, 2010; Pach-Gurgul, 2012). Therefore, by emphasising the willingness to research energy cul­ tures, it is quantitative data that are often analysed, and their results are extrapolated to public awareness and social practice, while no attempt is made at a broader analysis of awareness with the aid of qualitative techniques. This however does not mean that no research is conducted into awareness o f - for instance - the practice adopted by energy consumers, citizens’ attitude to particular energy sources, etc. Qualitative research of the “man-energy” kind abounds, but not all o f it follows from the willingness to analyse the problem o f energy culture. Frequently, the research is concerned with the analysis of environmental awareness, environment-friendly practice or attitudes to infrastruc­ ture investments, such as nuclear power plants or wind farms (Cf. Papuziński, 2006: 33-40; Tuszyńska, 2007: 233-236; Hłobił, 2010: 87-94).

Given the necessity to make the research problem more specific, the text addresses the following research questions: (1) Is the claim that the European Union presents spe­

cial "energy cultures” legitimate?, (2) Did the decade o f2001-2011 witness changes in the fie ld o f the European Union "energy cultures”, as earlier recognized by the litera­ ture? In order to see the research process through, the following working research hy­

potheses have been subjected to verification: (1) It must be posited that the

dissimilarities in energy structures ofparticular E U member states are an outstanding premise on which to base a recognition o f the existence o f specific ‘energy cultures’in the European Union, (2) It must be assumed that in the period o f 2001-2011 there were

226 Remigiusz ROSICKI

changes in the area o f ‘energy cultures’ o f the European Union, which had earlier been recognised in the literature.

THE CONCEPT AND EXAMPLES OF ‘ENERGY CULTURES’

Z. Łucki and W. Misiak define an energy culture as “a humanistic strategy in the ap­ proach to energy issues, both on the part of producers and consumers, which enables a resolution o f social problems.” Interestingly, they set out to analyse the term of an en­ ergy culture by referring to The Two Cultures, a well-known text on the history of sci­ ence by C. P. Snow. Snow’s text serves as a pretext for invoking the bipolar division of sciences on account of the methodological dimension into idiographic and nomological ones. The authors point out that their wish is to approach an energy culture as a field of research, through three prisms - the way Snow characterised scientific cultures - (“sci­ entific”, “humanistic”, and “social”). For Z. Łucki and W. Misiak the following consti­ tute component parts of an energy culture: (1) decarbonisation o f human activity, (2) an improvement in the efficient use o f energy carriers, (3) proper regard for energy dis­ played by consumers, (4) supplying the public with cheap and reliable energy (Snow, 2001; Łucki, Misiak, 2010: 47-49).

Both the authors point to two main sets o f factors influencing an energy culture: (1) the economic one, and (2) the social one. The same factors are invoked by A. Pach-Gurgul, however she erroneously quotes - after J. L. Campbell - other factors influencing an energy culture, e.g. the impact of the global culture, political trends. The reference is erroneous in view of the fact that by no means does J. L. Campbell’s text address the issues of an “energy culture”, while it analyses various ways of making pol­ itics in general (Pach-Gurgul, 2012:164; Campbell, 2002:21-38). Obviously, this does not mean that the factors p er se are not worth taking into account in the process o f ana­ lysing the set of factors influencing an “energy culture”.

In 2007 Tapio and his research team presented an analysis o f the decarbonisation processes applied to the economies of the European Union member states (EU-15) in the years 1970-2000. This research is invoked in the analyses by Z. Łucki and W. Misiak, as well as A. Pach-Gurgul, who reconstruct the types of energy cultures. In order to analyse the changes, Tapio and his colleagues employed three metrics: (1) de­ carbonisation of the whole economy, (2) decarbonisation of energy production and transport, and (3) dematerialisation o f the economy. For this purpose they performed a cluster analysis and generated six clusters: (1) Belgium, Germany and Great Britain

(Mature North European Economies 1), (2) France and the Netherlands (Mature North European Economies 2), (3) Sweden and Finland (Mature North European Economies

3), (4) Austria, Italy and Spain (Central and South European Economies), (5) Greece and Portugal (Peripheral South European Economies), (6) Denmark, Ireland and Lux­ embourg (as outlying countries) (Tapio et. al., 2007: 433-451). Despite the six clusters being distinguished in the research of Tapio and his team, it is possible to point to two kinds of culture: the North European and the South European one. It is worth emphasising this fact on account of the research presented in this text. In their book Z. Łucki and W. Misiak perform a synthesis of publications dedicated to “energy cul­

tures”, distinguishing nine cultures: (1) the Anglo-Saxon one, (2) the French one, (3) the Scandinavian one, (4) the Mediterranean one, (5) the East European one, (6) the Latin American one, (7) the Japanese one, (8) the Chinese one, and (9) the primitive one (Łucki, Misiak, 2010: 75-76). For instance, the “energy culture ofthe French type” is to be characterised by the predominance of nuclear power, a minimum supply of coal, a low emission and energy intensity ofthe economy, while the “energy culture ofthe Scandina­ vian type” is to be characterised by a great significance of renewable energy sources, a low consumption of coal, low-emission and intensity indexes of the economy. It is to be noted that in their typology, Łucki and Misiak included a description of energy markets, that is an attitude of individual states to sector liberalisation, which has been at length ad­ dressed in A. Midttun’s publications (Midttun, 2001; Finon, Midttun, 2004).

METHODOLOGY

The text sets out to verify the results of the research into the existence o f “energy cultures” in the European Union, which have been presented by the following persons: (1) A. Pach-Gurgul, (2) Z. Łucki and W. Misiak (3) P. Tapio, D. Banister, J. Luukkanen, J. Vehma and R. Willamo. As regards A. Pach-Gurgul, and P. Tapio and his research team, a statistical analysis was carried out on the basis of the following research tools employed for the purpose o f the cluster analysis: Ward’s method, complete linkage, k-means, multidimensional scaling. A comparative analysis of the research processes authored by the above-mentioned researchers is difficult because of: (1) a selection of various statistical instruments (e.g. for the purpose o f cluster analysis A. Pach-Gurgul uses Ward’s method, the complete linkage algorithm, the k-means method and multidi­ mensional scaling, whereas P. Tapio and his research team use the complete linkage al­ gorithm, and Z. Łucki and W. Misiak do not conduct a statistical analysis at all, limiting themselves only to a synthesis and referring to other scientists’research), (2) a selection of various indexes characterising potential and distinguished “energy cultures” (Tapio et. al., 2007: 433-451; Łucki, Misiak, 2010: 43-91; Pach-Gurgul, 2012: 159-177).

The object of the present text is not to discuss research tools or to establish which one of the statistical methods is more effective - the complete linkage algorithm or perhaps other methods of cluster analysis. Such problems are usually analysed along with a presentation of various statistical methods (Cf. Aldenderfer, Blashfield, 1984; Kaufman, Rousseeuw, 2005; Stanisz, 2007; Everitt, Landau, Leese, Stahl, 2011; Mirkin, 2015). The same attitude was adopted in relation to the differences in the selection of indexes characterising “energy cultures”. It is noteworthy that the reasons for using particular indexes concerned with the use of energy may be, for instance, pragmatics-motivated (an access to data) or teleol­ ogy-motivated (related to the objective that has been set in a research process).

However, it must be pointed out that on account o f the willingness to conduct a com­ parative analysis, following A. Pach-Gurgul, only a selection of statistical methods have been adopted for the purpose ofthe research, to wit: (1) Ward’s method (the mini­ mum variance method), (2) the k-means method (Pach-Gurgul, 2012: 159-177). The same applied to the indexes used for the purpose of classification of the clusters, which in turn could be related to particular “energy cultures”. The problem o f similarities and dissimilarities in this scope will be discussed in the subsequent point of the text.

228 Remigiusz ROSICKI

Also, it must be emphasised that in their analyses, both P. Tapio and his research team, and A. Pach-Gurgul subjected a different number of states to analysis, and adopted different timeframes. As for P. Tapio and his research team, 15 member states of the so-called “old” European Union (EU-15) were analysed within the period of 1970-2000; A. Pach-Gurgul’s analysis covered 27 European Union member states (EU-27) and the focus was 2008. In the present text the subjective scope of the analysis covers 28 European Union member states (EU-28) and the indexes o f 2011. Further­ more, for comparative purposes the same kind o f analysis of 2001 is employed. Note­ worthily, the text does not include detailed calculations in this respect.

Some light should be shed on the methods used to verify the hypotheses and the comparative analysis. As mentioned before, the research used clustering, which is a common method for analysing empirical data. The method serves to group objects and variables in indefinite groups.

In the first place - following A. Pach-Gurgul - the use was made of Ward’s method, which is one o f the most frequently employed agglomerative clustering method. The characteristic feature of this method is the use o f a variance analysis for the purpose of determining the distance between clusters. The distance between one cluster composed of objects and another one cannot be directly expressed by way of the distance between the objects belonging to these clusters (Kaufman, Rousseeuw, 2005: 230-234; Stanisz, 2007: 122; Mirkin, 2015: 111-136). Hence, “the method aims to minimise the sum of squared deviations of any two clusters which can be formed at any stage” (Analiza

skupień, 2015). Therefore, clusters that “ensure the minimum sum of squared distances

from the centre o f mass o f a new cluster, which they create” are merged (Rosz- ko-Wójtowicz, 2014: 74). The literature points out that this kind o f agglomerative method is cognitively effective, however it yields small, and yet most natural clusters (Roszko-Wójtowicz, 2014: 74; Analiza skupień, 2015).

Another method employed is the method o f k-means (i.e. a non-hierarchical algo­ rithm o f cluster analysis). This clustering method is in its design different from agglomerative and divisive methods (including Ward’s method). Hierarchical methods generate arranged cluster trees, whereby lower-order clusters are subsumed under higher-order ones; the k-means method divides clusters in such a manner that no cluster is a sub-cluster of another cluster (Stanisz, 2007: 127-128). The choice of a specified number of clusters results in groups of objects that are most similar (close), whereas ob­ jects from other groups are most different (distant). Noteworthily, it must be pointed out that it is the analysing person who makes an arbitrary choice as to the number of groups made up of particular objects (Sokołowski, 2002; Mirkin, 2015: 75-110).

INDEXES OF ‘ENERGY CULTURES’

A selection of indexes

The introduction contains assumptions to be verified, that is the assumptions of the existence or non-existence o f special “energy cultures”, as well as the assumptions of changes occurring in the energy cultures earlier recognised in the literature. A selection

of indexes characterising an “energy culture” is essential for this purpose. Arguably, the very choice of indexes is a definiens o f sorts of the term in question. The text refers to indexes chosen by Pach-Gurgul, but given the lack of vital data concerned with the manner of calculating some indexes, the text does not undertake an appropriate com­ parative analysis. The above-mentioned author employed the following indexes for the purpose of the analysis of the EU-27 “energy culture” : (1) an index of energy intensity (kgoe/€1000), (2) an index of energy consumption per capita (toe/cap), (3) an index of GHG emission per capita (t CO2 equiv.), (4) a level of primary energy production by fuel kinds in Mtoe (coal, petroleum, gas, nuclear fuel, renewable energy sources) (Pach-Gurgul, 2012: 165-177).

In the text, given the difficulty o f characterising “energy cultures” of the European Union member states, the following indexes have been adopted: (1) the energy intensity of an economy (kgoe/€1000), (2) energy consumption per capita (toe/cap), (3) the emission o f greenhouse gases p er capita (t CO2 equiv.), (4) a level o f gross inland en­ ergy consumption in Mtoe (for solid fuels, petroleum and petroleum products, gas, nu­ clear fuel, renewable energy sources, non-renewable energy sources). The selection of indexes shows a visible difference in that the index of primary energy production has been replaced with the index of gross inland energy consumption with a different fuel structure (Tab. 1).

Table 1 Indexes of energy cultures in the European Union member states in 2011 (EU-28)

S T A T E S Energy intensity of the economy Gross in­ land con­ sumption of energy p e r capita Green­ house gas emis­ sions per capita

Gross inland energy consumption

Solid fuels Petro­ leum and p. products Gas Nucle­ ar fuel RES -S on ES n R e 0) o 0 g 0 1 -w ^ (toe/cap) O2 iv.) C ui

tq(e (Mtoe) (Mtoe) (Mtoe) (Mtoe) (Mtoe) (Mtoe)

[1] [2] [3] [4] [5] [6] [7] [8] [9] BE 182.1 5.43 13.43295 3.4 22.9 15.1 12.4 2.9 2.7 BG 7G5.5 2.62 7.41G16 8.1 3.7 2.6 4.2 1.4 G.G CZ 355.4 4.13 9.25921 18.4 9.G 6.8 7.3 3.2 G.2 DK 89.7 3.42 8.19344 2.5 7.1 3.5 G.G 4.2 G.4 DE 129.2 3.87 9.62759 8G.4 1G8.3 69.6 27.9 29.7 3.9 EE 5G5.8 4.60 4.31386 4.1 1.1 G.5 G.G G.8 G.G IE 83.7 3.03 7.75152 2.G 7.G 4.1 G.G G.8 G.G EL 154.4 2.47 8.55788 7.9 13.5 4.G G.G 2.1 G.G ES 135.1 2.79 6.8G64G 12.3 57.7 28.9 14.9 14.7 G.2 FR 142.6 3.99 5.73271 1G.2 82.G 37.G 114.1 18.1 1.2 HR 231.9 1.94 3.97683 G.7 3.7 2.6 G.G G.9 G.G IT 12G.7 2.85 6.745G6 15.9 67.3 63.8 G.G 2G.G 1.1 CY 174.G 3.18 7.792G5 G.G 2.6 G.G G.G G.1 G.G LV 333.5 2.05 3.65131 G.1 1.4 1.3 G.G 1.4 G.1

230 Rem igiusz RO SICK I [1] [2] [3] [4] [5] [6] [7] [8] [9] LT 298.7 2.32 4.29685 0.2 2.4 2.7 0.0 1.1 0.0 LU 136.7 8.96 23.10506 0.1 2.9 1.0 0.0 0.1 0.0 HU 281.6 2.53 5.22968 2.8 6.4 9.4 4.1 1.9 0.1 MT 163.5 2.71 18.41715 0.0 0.9 0.0 0.0 0.0 0.0 NL 144.7 4.88 14.39552 7.5 32.5 34.3 1.1 3.3 0.8 AT 125.0 4.04 8.43828 3.5 12.1 7.8 0.0 8.8 0.7 PL 314.7 2.65 8.01486 54.6 25.7 12.8 0.0 8.0 0.6 PT 150.8 2.26 4.94078 2.2 11.4 4.5 0.0 5.1 0.2 RO 393.7 1.70 4.08505 8.2 9.3 11.1 3.0 5.1 0.0 SI 230.6 3.54 7.99669 1.5 2.6 0.7 1.6 1.0 0.0 SK 349.3 3.23 6.39849 3.7 3.6 4.6 4.0 1.3 0.0 FI 210.4 6.65 9.81307 5.7 10.0 3.4 6.0 9.1 0.2 SE 149.4 5.26 5.94618 2.5 14.0 1.2 15.6 16.5 0.5 UK 102.8 3.18 7.78824 30.6 69.7 70.2 17.8 7.7 0.7

Source: Author’s own study based on Eurostat, IEA data.

A description of indexes

The first index is energy intensity, which is concerned with the operation o f an econ­ omy. It defines energy consumption in production processes in particular sectors of econ­ omy. The differences in the level of energy intensity may be indicative of the level of economic development, an economic structure, a technological advancement, the type of carriers used in the primary energy production. A low level of energy intensity points to a higher efficiency of energy sources management. It must however be pointed out that a low energy intensity may characterise economically underdeveloped countries, with the provision that we are dealing with a low energy consumption in general. As regards en­ ergy intensity, there is a variety o f indexes to choose from: (1) an index of GDP energy in­ tensity (calculated as the ratio of the amount of energy consumed to the value of GDP), (2) an index of energy consumption per capita (calculated as the ratio of primary energy consumed to the population) (Pach-Gurgul, 2012: 166-167; Energy glossary, 2015).

Another index is the greenhouse gas emission calculated as the ratio of the emission to the population (Greenhouse gas emission statistics, 2015). Emissions may be sug­ gestive of an energy structure of particular countries, a technological advancement and the level o f clean energy sources development. Taking into account emissions of very small countries, such as Luxembourg, may become problematic for analyses. The ten­ dency which should be included in the emission analyses presupposes a correlation be­ tween a GDP growth and an emission increase. The index level may also be affected by the dominance of particular carriers in the energy production structure.

The last index to characterise an “energy culture” is the gross inland energy consump­ tion expressed in Mtoe. Unlike the selection of indexes made by A. Pach-Gurgul, the text does not employ the index based on primary energy production by the type of fuel. The index of gross inland energy consumption takes into account the total energy demand in a given country, and as such it defines the amount of energy required to meet the con­

sumption needs. This kind of consumption embraces: (1) energy consumption by the en­ ergy industry, (2) losses sustained in relation to distribution and processing, (3) the ultimate energy consumption by end users (Eurostat statistics explained, 2015). The ele­ ments taken into account while calculating the index of gross inland energy consumption are: (1) primary production, (2) the so-called reclaimed products, (3) net energy imports, (4) changes in stores (however, not including the so-called energy/oil stored in interna­ tional sea bunkers) (Energy glossary, 2015; Eurostat statistics explained, 2015).

AN ATTEMPT AT GROUPING THE EUROPEAN UNION STATES Ward’s method

Ward’s method was employed to group the European Union member states (EU-28) in 2011, which made it possible to distinguish four existing clusters: (1) Belgium, the Netherlands, Austria, Sweden, Finland and Luxembourg (however Luxembourg stands out against the other countries); (2) Bulgaria, the Czech Republic, Estonia, Poland, Croatia, Latvia, Lithuania, Hungary, Slovakia and Romania (a clear division o f these countries into the former four and the latter six is observable); (3) Denmark, Ireland, Greece, Portugal, Cyprus, Slovenia and Malta (however Malta stands out against the other countries); (4) Germany, Spain, Italy, Great Britain and France, however Ger­ many stands out against the other countries (Diag. 1).

Diagram 1. A dendrogram (Ward’s method with Euclidean distance) BE NL AT SE FI LU BG CZ EE PL HR LV LT HU SK RO DK IE EL PT CY SI MT DE ES IT UK FR 0 1 2 3 4 5 6 Linkage Distance

1

---— ■!

---5h

=>1

---—

h >

232 Rem igiusz RO SICK I

The A-means method

With the optimal number o f clusters established, the text sets out to find out what ar­ eas characterising an “energy culture” display the biggest differences between the dis­ tinguished groups (Tab. 2). Thanks to the k-means method it is possible to conclude that the differences between distinguished clusters are highly significant (p<0.01) in respect of almost all the analysed variables, with the exception of nuclear fuels, where the dif­ ferences are significant with the significance level of 0.05 (Tab. 3).

Table 2 The mean values of the analysed indexes in individual clusters (A:-means method)

Individual indexes: Cluster

1 2 3 4 5 6

Energy intensity o f the economy G.G669 G.6218 G.1121 G.334G G.G732 G.1139 Gross inland consumption o f energy

per capita

G.2G7G G.2479 G.29G8 G.G813 G.2989 G.65G6

Greenhouse gas emissions per capita G.16G2 G.185G G.2858 G.G516 G.3G72 G.685G

Solid fuels G.2146 G.2649 G.G354 G.G318 1.GGGG G.G456

Petroleum and petroleum products G.6357 G.G836 G.G638 G.G424 1.GGGG G.1726

Gas G.7119 G.G8G8 G.G391 G.G737 G.9915 G.2393

Nuclear fuel G.3216 G.G252 G.G226 G.G139 G.2445 G.G394

RES G.5G93 G.1128 G.1594 G.G8G8 1.GGGG G.G7G7

bo b-R E S G.2G51 G.G513 G.G513 G.G147 1.GGGG G.2991

Source: Author’s own study.

Table 3 Cluster variations - variance analysis results (ANOVA)

Individual indexes: ANOVA Between groups variance df Within groups variance df F Test P Value

Energy intensity o f the economy 0.99 5 0.40 22 10.85 <0.0001

Gross inland consumption o f energy per capita

0.70 5 0.56 22 S.48 0.0020

Greenhouse gas emissions per capita 0.90 5 0.48 22 8.31 0.0002

Solid fuels 1.03 5 0.30 22 14.83 <0.0001

Petroleum and petroleum products 1.79 5 0.12 22 64.86 <0.0001

Gas 2.08 5 0.40 22 22.63 <0.0001

Nuclear fuel 0.34 5 0.66 22 2.26 0.0846

RES 1.22 5 0.46 22 11.77 <0.0001

bo b-R E S 1.04 5 0.35 22 13.05 <0.0001

An extension of Table 2 is Diagram 2, which graphically presents mean levels of analysed features for the distinguished clusters. The first cluster includes Spain, France, Italy and Great Britain (which is motivated by relatively high scales of gross energy con­ sumption). The second cluster includes Bulgaria, the Czech Republic, Estonia and Po­ land (the characteristic feature here is the highest level of energy intensity - kgoe/€1000). The third cluster includes Denmark, Ireland, Greece, Cyprus, Malta, Austria, Slovenia, Finland and Sweden. This group is characterised by low results in all the scales, the high­ est ones concerning the index of energy consumption per capita (toe/cap) as well as the index of GHG emission per capita in relation to energy consumption (t CO2). The fourth cluster includes Croatia, Latvia, Lithuania, Hungary, Portugal, Romania and Slovakia. This group o f countries is characterised by the lowest results in all the scales, except for energy intensity (kgoe/€1000). The fifth cluster embraces only Germany, characterised by high scales of gross inland energy consumption. The last, sixth cluster contains the following countries: Belgium, Luxembourg and the Netherlands. This group is character­ ised by the highest indexes of energy consumption per capita (toe/capita), as well as the index of GHG emission per capita in relation to energy consumption (t CO2).

Diagram 2. Means of particular clusters (A-means method)

—o - Cluster 1 - o - Cluster 2 - o ■ Cluster 3 _a— Cluster 4 ♦ Cluster 5 C lu ster6

Individual indexes: [1] Energy intensity of the economy, [2] Gross inland consumption of energy p er capita,

[3] Greenhouse gas emissionpercapita, [4] Solid fuels, [5] Petroleum and petroleum products, [6] Gas, [7] Nucle­ ar fuel, [8] RES, [9] mom-RES.

Source: Author’s own study.

The main object o f the analysis presented in the text was to point to and confirm the existence o f special “energy cultures” in the European Union. In order to achieve this

234 Rem igiusz RO SICK I

aim the use was made o f research present in the literature, inter alia, in the publications containing statistical analyses by: (1) A. Pach-Gurgul, (2) P. Tapio and his research team - Banister, J. Luukkanen, J. Vehma i R. Willamo, also in a review, (3) Z. Łucki and W. Misiak. Compared with the invoked research, the timeframe of the analysis in the text covered 2011, and the subjective scope embraced 28 countries (EU-28).

As part of the research process the following research hypotheses were subjected to verification: (1) It must be assumed that the dissimilarities in energy structures o f par­

ticular E U member states are an outstanding premise on which to base a recognition o f the existence o f specific "energy cultures” in the European Union, (2) It must be as­ sumed that in the period o f 2001-2011 there occurred changes in the area o f "energy cultures” o f the European Union, which had earlier been recognised in the literature.

In this scope the following conclusions ought to be presented:

I. CONCLUSIONS: The differences in energy structures result in the possibility of pointing to particular features characterising member states and clusters of countries, partially because of the predominance o f specific energy carriers and/or a way o f their conversion. On account of the use of selected indexes for the purpose of characterising the specificity o f the European Union member states, six clusters have been distin­ guished, and they include the following countries: (1) Spain, France, Italy and Great Britain; (2) Bulgaria, the Czech Republic, Estonia and Poland; (3) Denmark, Ireland, Greece, Cyprus, Malta, Austria, Slovenia, Finland and Sweden; (4) Croatia, Latvia, Lithuania, Hungary, Portugal, Romania and Slovakia; (5) Germany; (6) Belgium, Lux­ embourg and the Netherlands.

An observable cluster that recurs in the research by A. Pach-Gurgul as well as in the analysis presented in the text covering 2011, embraces Bulgaria, the Czech Republic, Estonia and Poland (Cluster II). As for Ward’s method the cluster is broader, but fea­ tures an outstanding group of the above-mentioned countries (also including: Croatia, Latvia, Lithuania, Hungary, Slovakia and Romania). Besides, in this model, in compar­ ison with A. Pach-Gurgul’s research, another cluster is by and large repeated, including Croatia, Latvia, Lithuania, Hungary, Portugal, Romania and Slovakia - Cluster IV (for a variety o f reasons Croatia and Portugal are not repeated). The characteristic feature of both clusters is a higher scale o f the index of energy intensity (kgoe/€1000). This shows that it is legitimate to recognise the existence of the division of the European Union into the two “energy cultures” : the Western European one and Eastern European one. The main factor behind this division is the energy intensity of the economies of individual member states. Unlike the analysis by P. Tapio and his research team, the axis of divi­ sion has changed considerably, for within the so-called “old” European Union (EU-15) there was an observable general division into the Northern European and Southern Eu­ ropean ‘energy cultures’.

Both the above-mentioned clusters (II and IV) can be associated with the “Eastern European culture” as part of the division presented by Z. Łucki and W. Misiak. The en­ ergy culture distinguished by the two researchers is to be characterised by: (1) a variety of energy policies, (2) a high energy intensity of the economy, (3) a low electric energy consumption, (4) limited awareness, (5) a negative attitude towards energy market liberalisation. In the case of Bulgaria, the Czech Republic and Poland (Cluster II), the characteristic feature is a considerable share of solid fuels in the energy mix of the gross

inland energy consumption. The countries included in Cluster IV are characterised by a smaller share of solid fuels and a higher diversity of energy structure.

As part o f the k-means clustering method, it is worth pointing out two clusters of countries: Cluster I (Spain, France, Italy, Great Britain) and Cluster V (Germany). The countries in both clusters belong to the so-called “old” European Union, and their main feature is a substantial level of energy production, Germany being the largest producer. In both cases, the characteristic features are a marked share o f renewable energy sources, gas as well as petroleum and petroleum products in the gross inland energy consumption. Furthermore, both clusters are characterised by a low or relatively low in­ dex of energy intensity (kgoe/€1000).

As part o f the clustering based on the k-means method, in both periods, that is in 2001 and 2011 the cluster comprising Belgium, Luxembourg and the Netherlands was repeated. This cluster is characterised by a high index of energy consumption per capita and a high index of GHG emission per capita.

II. CONCLUSIONS: It must be emphasised that the period 2001-2010 is con­ nected with the transition process o f the electric energy and gas sectors in the European Union. The process is also associated with the introduction of the 1st Climate-Energy

Package in the European Union. What is more, the changes will be continued on ac­

count of the efforts at building an Energy Union as well as the implementation of the so-called 2nd Climate-Energy Package. O f particular significance is also the increase in the import dependence of the European Union, given the declining reserves of energy resources. Moreover, the objectives o f the “green energy policy” will be giving rise to the increasing energy efficiency and the growing share of renewable energy sources in energy production, which may constitute a factor limiting the import dependence. Thus, it should be posited that in the long term we will be witnessing a homogenisation of indexes, e.g. the index o f energy intensity.

These processes may result in the tendency for changes towards the model o f the ‘energy culture’, which Z. Łucki and W. Misiak have termed a “Scandinavian culture”, that is one characterised by the following features: (1) a development of RES, (2) a low consumption of coal, (3) a low index o f GHG emission, (4) a considerable energy con­ sumption, (5) a low energy intensity, (6) a heightened environment- and energy-con- sciousness.

By way of comparing the analyses for 2011 and for 2001, based on the k-means clustering method, it must be stressed that Belgium, Luxembourg and the Netherlands hold steady in the cluster (Cluster VI), whereby they abide by the roles played by the features earlier ascribed to them. A measure of stability can also be observed in the case of Spain, France, Italy, Great Britain and Germany (Clusters I and V). Noteworthily, Sweden has been eliminated from the group (despite lowering the scale o f the first three indexes).

The distinctive feature of Eastern European countries is the high scale o f the energy intensity index, as viewed against the backdrop of Clusters I and IV. Yet, some tendency towards stirring of the scale o f energy intensity index is to be observed in this scope. A transformation of the energy structure will also prove quite a challenge here. Poland can serve as an example, whereby coal will continue to be the country’s main electric energy carrier for the next 30 years.

236 Rem igiusz RO SICK I

The comparative analysis featuring in the text is limited, and so it does not address all the issues concerned with the correlations between the presented ‘energy cultures’. The work sets out to verify the results o f the research into the existence of ‘energy cul­ tures’ in the European Union, as presented by A. Pach-Gurgul, Z. Łucki and W. Misiak as well as P. Tapio and his research team. Still, the greatest emphasis has been laid on the proposals made by A. Pach-Gurgul, for it was her publication as well as the selec­ tion o f indexes and research tools included therein that proved to be an inspiration for the present text. It should also be noted that the object of analysis in the text does not embrace a discussion o f the legitimacy o f employing specific methods for clustering.

BIBLIOGRAPHY

Aldenderfer M. S., Blashfield R. K. (1984), Cluster Analysis, Newbury Park-London-New Delhi. Analiza skupień (2015), http://www.statsoft.pl/textbook/stathome_stat.html7http%3A%2F%2Fwww.stats-

oft.pl%2Ftextbook%2Fstcluan.html (01.07.2015). Burszta W. J. (1998), Antropologia kultury, Poznań.

Campbell J. L. (2002), Ideas, Politics, and Public Policy, “Annual Review of Sociology”, Vol. 28. Energy glossary, http://ec.europa.eu/eurostat/statistics-explained/index.php/Category:Energy_glos-

sary (01.07.2015).

Eurostat statistics explained, http://ec.europa.eu/eurostat/statistics-explained/mdex.php/Mam_Page (01.07.2015).

Everitt B. S., Landau S., Leese M., Stahl D. (2011), Cluster Analysis, London.

Finon D., Midttun A. (ed.) (2004), Reshaping European Gas and Electricity Industries Regulation, Markets and Business Strategies, London, New York.

Greenhouse gas emission statistics, http://ec.europa.eu/eurostat/statistics-explained/index.php/ Greenhouse_gas_emission_statistics (02.07.2015).

Hłobił A. (2010), Teoria i praktyka edukacji ekologicznej na rzecz zrównoważonego rozwoju, “Problemy Ekorozwoju”, No. 2, Vol. 5.

Kaufman L., Rousseeuw P. J. (2005), Finding Groups in Data An Introduction to Cluster Analysis, New Jersey.

Kłosowska A. (1969), Z historii i socjologii kultury, Warszawa. Kłosowska A. (1972), Społeczne ramy kultury, Warszawa.

Łucki Z., Misiak W. (2012), Energetyka a społeczeństwo. Aspekty socjologiczne, Warszawa. Midttun A. (2001), Perspectives on Commercial Positioning in the Deregulated European Electricity

Markets, in: European Energy Industry Business Strategies, (ed.) A. Midttun, London-New York.

Midttun A. (ed.) (2001), European Energy Industry Business Strategies, London-New York. MirkinB. (2005), Clustering fo r Data Mining A Data Recovery Approach, Boca Raton-London-New

York-Singapore.

Pach-Gurgul A. (2012), Jednolity rynek energii elektrycznej w Unii Europejskiej w kontekście bezpieczeństwa energetycznego Polski, Warszawa.

Papuziński A. (2006), Świadomość ekologiczna w świetle teorii i praktyki. Zarys politologicznego modelu świadomości ekologicznej, “Problemy Ekorozwoju”, No. 1, Vol. 1.

Roszko-Wójtowicz E. (2014), Analiza skupień w ocenie warunków pracy w krajach Unii Euro­ pejskiej, “Wiadomości Statystyczne”, No. 11.

Snow Ch. P. (2001), The Two Cultures, London.

Sokołowski A. (2002), Metody stosowane w data mining, http://www.statsoft.pl/Portals/0/Down- loads/Metody_stosowane_w_data_mining.pdf (30.07.2015).

Stanisz A. (2007), Przystępny kurs statystyki z zastosowaniem STATISTICA PL na przykładach z medycyny, t. 3: Analizy wielowymiarowe, Kraków.

Strinati D. (1998), Wprowadzenie do kultury popularnej, Poznań.

Tapio P., Banister D., Luukkanen J., Vehmas J., Willamo R. (2007), Energy and transport in compari­ son: Immaterialisation, dematerialisation and decarbonisation in the EU15 between 1970 and 2000, “Energy Policy”, Vol. 35, Issue 1.

Tuszyńska L. (2007), Świadomość środowiskowa społeczeństwa polskiego, “Problemy ekologii”, No. 5.

ABSTRACT

The main objective o f the text is to present an analysis that points to the existence o f special “energy cultures” in the European Union. The comparative analysis encompassing the results of previous research into “energy cultures” employs statistical methods, i.e. a cluster analysis (Ward’s clustering method and k-means clustering method). The main sections o f the text ad­ dress: (1) the concept and examples of “energy cultures”, (2) a methodology of analysis, (3) ase- lection o f indexes characterising “energy cultures”, (4) an attempt at grouping the European Union member states with the aid of clustering, (5) conclusions. With a view to making the re­ search problem more specific, the present text features the following questions: (1) Is the claim that the European Union manifests special ‘energy cultures’ legitimate?, (2) D id the decade o f 2001-2011 witness changes in the fie ld o f the European Union ‘energy cultures’, as earlier re­ cognised by the literature?

Key words: energy cultures, indexes o f energy cultures, energy policy, European Union

KULTURY ENERG ETYCZNE W UNII EUROPEJSKIEJ STRESZCZENIE

Głównym celem tekstu jest prezentacja analizy, która wskazuje na istnienie szczególnych “kultur energetycznych” w Unii Europejskiej. W analizie porównawczej z dotychczasowymi wynikami badań nad “kulturami energetycznymi” wykorzystano metody statystyczne, tj. anali­ zę skupień (grupowanie metodą Warda i grupowanie metodą k-średnich). Główne części pracy dotyczą: (1) pojęcia i przykładów “kultur energetycznych”, (2) metodologii analizy, (3) doboru wskaźników charakteryzujących “kultury energetyczne”, (4) próby grupowania państw człon­ kowskich Unii Europejskich za pomocą analizy skupień, (5) wniosków. W celu uszczegółowie­ nia problemu badawczego w pracy przedstawiono następujące pytania: (1) Czy zasadne jest twierdzenie, że w Unii Europejskiej mamy do czynienia ze specyficznym i ‘kulturami energetycz­ n y m i’?, (2) Czy w okresie 2001-2011, następują zm iany w obrębie, stwierdzonych wcześniej w literaturze tematu, ‘kultur energetycznych’w Unii Europejskiej?

Słowa kluczowe: kultury energetyczne, wskaźniki kultur energetycznych, polityka energetycz­